Sohini Roychowdhury sohiniroych

I am an Expert in Applied ML and AI applications specializing in computer vision and image processing tasks.

sohiniroych / adjust_data.py

Created March 10, 2021 23:34

generate binary mask from grayscale images

	def adjustData(img,mask,flag_multi_class,n_class):

	if(flag_multi_class):
	img /= 255
	mask = mask[:,:,:,0] if(len(mask.shape) == 4) else mask[:,:,0]
	new_mask = np.zeros(mask.shape + (n_class,))
	for i in range(n_class):

	new_mask[mask == i,i] = 1
	new_mask = np.reshape(new_mask,(new_mask.shape[0],new_mask.shape[1]new_mask.shape[2],new_mask.shape[3])) if flag_multi_class else np.reshape(new_mask,(new_mask.shape[0]new_mask.shape[1],new_mask.shape[2]))

sohiniroych / loss_functions.py

Created March 10, 2021 23:31

loss functions for semantic segmentation

	def dice_coef(y_true, y_pred, smooth=1):
	intersection = keras.sum(y_true * y_pred, axis=[1,2,3])
	union = keras.sum(y_true, axis=[1,2,3]) + keras.sum(y_pred, axis=[1,2,3])
	return keras.mean( (2. * intersection + smooth) / (union + smooth), axis=0)

	def dice_coef_loss(y_true, y_pred):
	return -dice_coef(y_true, y_pred)

sohiniroych / data_prep.py

Created March 10, 2021 23:29

Preparing 3 class data

	def prepare_multi_class_GT(GT_PATH, class_names, savepath, target_size=(512,512), n_class=3):

	f_names = os.listdir(GT_PATH+class_names[0])
	for files in f_names:
	GT_im=np.zeros(np.concatenate((target_size,n_class),axis=None)) #This creates a zero array of size (512,512,3)
	FG=np.zeros(target_size)

	for idx,cn in enumerate(class_names):
	lab=io.imread(GT_PATH+cn+files, as_gray=True)
	lab = trans.resize(lab,target_size)

sohiniroych / data_gen.py

Created March 10, 2021 23:27

	data_gen_args = dict(rotation_range=0.3,
	width_shift_range=0.1,
	height_shift_range=0.1,
	shear_range=0.1,
	zoom_range=[0.7,1],
	horizontal_flip=True,
	fill_mode='nearest')

sohiniroych / cloudbuild.yaml

Created February 17, 2021 21:45

cloudbuild.yaml

	steps:
	# Step 1: pull the container image if it is already built, if fails do not exit
	- name: 'gcr.io/cloud-builders/docker'
	entrypoint: 'bash'
	args:
	- '-c'
	- 'docker pull gcr.io/$PROJECT_ID/appcicd:latest \|\| exit 0'
	# Step 2: Create a docker image if none exists, esle load existing one
	- name: 'gcr.io/cloud-builders/docker'
	args:

sohiniroych / cloudbuild.yaml

Created February 17, 2021 21:42

YAML file

	steps:
	# pull the container image if it is already built
	- name: 'gcr.io/cloud-builders/docker'
	entrypoint: 'bash'
	args:
	- '-c'
	- 'docker pull gcr.io/$PROJECT_ID/appcicd:latest \|\| exit 0'
	# build the container image
	- name: 'gcr.io/cloud-builders/docker'
	args:

sohiniroych / Dockerfile

Created February 17, 2021 21:40

Dockerfile

	FROM python:3.6-slim-buster

	RUN python -m pip install --upgrade pip
	COPY requirements.txt .
	RUN pip install -r requirements.txt

	COPY . /app
	WORKDIR /app/app_files

	CMD ["python", "app.py"]

sohiniroych / app.py

Created February 17, 2021 21:37

	if __name__=='__main__':
	app.run( debug=True, host='0.0.0.0', port=int(os.environ.get('PORT', 8080)) )
	label="Regression_model", linewidth=2)

sohiniroych / model.py

Created February 17, 2021 21:31

Model Fitting and visualization

sohiniroych / model.py

Created February 17, 2021 21:24

	from sklearn.tree import DecisionTreeRegressor
	reg=DecisionTreeRegressor(max_depth=5)
	reg.fit(X,y)
	y_pred=reg.predict(X)
	print(reg.score)